Real-time notes from Locarno WG 4, Wednesday June 8 2005

Fletcher: introduction & preliminaries
  
How to handle Loukas's request for constraints - end each talk with a zinger
Smmarizing coronal-source results - a survey paper?
Summarizing footpoint motions? (Sakao '94)
Quiet-sun science for RHESSI

Dennis: Separating thermal & nonthermal sources

Need to use all techniques (temporal, spatial, spectral, line spectrum)
Difficulties with continuum (from Amir Caspi), eg 23-Jul 00:41-00:41:30
  spectral fits + image information
Phillips RESIK event, 26-Apr-2003: light curves from RHESSI
Photon spectrum for spike shows clear separation between thermal and
  nonthermal components (Fletcher: what about intermediate components?
  Pick: are the footpoints resolved? Christe: how to separate in images?
  Schmahl: how about CLEAN boxes?)
Counts spectrum of same hopelessly confusing; OSPEX is doing too good a
  job that must not be blindly trusted! Note an unexplained low-energy
  excess (<5 keV) that is definitely not thermal, but images to looptop)
Two-temperature fit for a different time interval
Feature ratio temperature (Caspi) for several flares (ratio temperature
  vs continuum temperature) show patterns resembling Phillips calculations.
  July 23 early phase is distinguishably different
Emissivity vs temperature (eq. width equivalent) again shows systematic
  trends
April 21 decay phase (Phillips) equivalent-width in the range 0.5-2 keV
  show clearly non-photospheric abundances
RESIK flare EW measurements are more scattered, but A1 points look
  pretty consistent except to too high a temperature
April 23 flare still different again, huge EW of 4-5 keV in flare
  decay, 25=>15 MK
Conclusions: have real features; centroid energies vary systematically;
  Fe/Fe-Ni varies mysteriously; EW varies but mysteriously; eagerly
  await XSM

McTiernan

Why RHESSI and GOES together?
  Data exist for all flares; we want to do all flares; have had trouble 
  putting data together in the past, but models have improved
DEM plot for a RHESSI flare can show a sharp peak (just RHESSI data)
XDEM software: multiple instruments, CHIANTI_KEV; RHESSI DRM, GOES_TRANSFER,
  TRACE_EUV_RESPONSE, SXIG12_LAMBDA_RESPONSE; inclusion of non-thermal
  component also possible in principle
RHESSI-GOES temperature response functions very broadband, D(resp)/DT smal
DEM calculation via PIXON inversion ("fuzzy pixons", parabolic in shape)
Test case 19-sep-02 decay phase
Quicklook spectra for this test case
DEM results from 2 GOES channels, about 25 RHESSI channels @2/3 keV
  resolution; start with power-law DEM and generate 0.5K temperature bins
Results! Error bars a bit schematic, but see 3 nice components
Movie of results shows low-T component disappearing; 12 and 25 keV
  components persist; peak positions do not vary
Fits and residuals: fit vector of GOES and RHESSI channels is not very
  good, with systematic residuals confirming that (CANDIDATE SLIDE FOR
  WORST_SLIDE AWARD)
Same with power-law input DEM also shows systematic biases (ANOTHER)
Case B: infer that broad response functions are leading to bad fits
Case C: counts spectrum from 20-Jul-02 flare; problems at highest and
  lowest energies => endpoint errors from PIXONS
Case C with DRM is better
Conclusions: some progress but have systematic problems; for future will
  try fractal pixons, larger RHESSI energy bins, SXI and TRACE input,
  and also nonthermal component
 
Vaananen

XSM (SMART-1) overview

Small, cheap Si non-imaging spectrometer, with resolution ~.33 keV (FWHM)
  at 6 keV, spectral range 1-20 keV, time resolution 16s, two units for
  albedo analysis of lunar soil under solar X-irradiation
In-flight Fe-56 calibration
Cross-reference between GOES and XSM fluxes
Saturation @ 200,000 counts (per 16 s?)
Sensitivity to 3.1 x 10^-8 W/m^2 (GOES units?)
March '04 to April '05 time range 

Christe

Past work: adopted from Markus
Flare list, take A1 < class < C1 and get 1500 total events
Neat map for the 5-month interval shows only ARs (Schmahl comment about
  finding microflares in the quiet Sun in RHESSI images - ?)
Max counts distribution for 6-12 keV shows nice power law at -1.4;
  durations peak at about 2 min w/ rolloff at about 1000 sec
Spectral fitting with Hannah's OSPEX code, nighttime backgrounds, 16-sec
  integrations => peak temperature at about 10 MK and a steep roll-off
  below 8-10 MK; power-law components soft (5-7)
Non-thermal energy distribution peaked at 10^29 ergs, but broad
Final comparison with Crosby et al. by using a 25 keV threshold shows about
  a factor of 10 difference - solar cycle?
Question marks: an em/T scatter plot shows no hot, low-em events; McTiernan
  and Dennis think of systematic ways this might happen
Neupert effect? Find some correlation between peak thermal counts and peak
  nonthermal counts. McTiernan recommends referring to Lee et al. (~1995)
  for treatment of limited parameter spaces and non-parametric tests
Gamma vs view angle: slight hardening at center
Conclusions: no evidence for a small-flare steep distribution; no evidence
  for enough energy to heat the corona (down to 10^29 ergs)

Schmahl

Preflare phenomena in HXR, SXR and EUV
Harrison preflare blips as seen by HXIS; Ohyama-Shibata; Sui-Holman 
  current sheet
3-nov-03 early rise; RHESSI counts heavily confused with particles
CLEAN and PIXON movies of preflare (to 01:08:30); SXI movies
  RHESSI 12 keV source clearly above the limb
EIT movie shows very large coronal loops
RHESSI preflare maps 01:07:30-1:12:30 show coronal source being extinguished
  by dynamic-range limitation
RHESSI spectrum of preflare phase barely shows Fe feature but A1 state; 
  (McTiernan QL spectra _do_ show Fe feature)
MLSO CME from 01:23 @ 50-100 km/s, starting at impulsive maximum?
Nobeyama nice images but no pre-flare source at "termination shock" (Pick
  comment about coherent radio sources at preflare coronal site in other
  events; Hudson comment that Shibata-cartoon termination shock would
  predict strong synchrotron emission))
Numerous questions remain... cf Kundu presentation

Gallagher (& Milligan)

Chromospheric evaporation
Hurford "napkin" cartoon
Worry about low-energy cutoff in HX spectrum
Two regimes of evaporation, "gentle" below F < 10^10 or so; "explosive"
26-Mar-02 CDS/TRACE example, RHESSI missed the peak - probably the 
  very best CDS flare observation
12-Jun-03? Not sure which event is being discussed below. Milligan's SPD
  event?
PIXON images overlying the CDS images
CDS velocity separation into two components as a function of temperature
  shows redshifts below 10^6 and upflow above
RHESSI spectrum shows steep powerlaw, -7.3, flux 2e10 @ area = 1e18
  (Hudson comment about UV footpoint areas implying a lower limit here)
Compare with the many 1D radiative hydro modeling results and find just
  at the upper limit of the "gentle" domain
Movie of CDS velocities for 26-Mar-02. Footpoints look synchronous. 
   Fletcher comment about Czaykowska-Alexander papers

Temmer

Footpoint analysis 17-Jan-05
Conjugate temporal behavior in four episodes and with total
Reconnection rate = B_corona*v, distances by fit to neutral-line position;
  separately for each footpoint (see Asai's paper for this method; several
  questions about what is correct or plausible)
M. Pick discussion of how to delay times between footpoints?
Find poor correlations between inferred reconnection rates and hard X-ray
  rates
H-alpha ribbon expansion movie, with RHESSI overlay (Fletcher comment about
  forcing a 2D model when 3D might be more instructive)
Discussion of degree of conjugacy, look forward to Fletcher presentation
Movie of H-alpha overlaid on MDI, with RHESSI
Nice movie of footpoints' regular motion across magnetogram
Cartoon suggesting inclination of field 
Summary - reconnection rates, B and v correlation, absence of FPs within
  umbrae and suggestion of tilted field lines
Ji question about inward motions...

Ji

H-alpha kernel motion and coronal sources
Centroid definitions and caution about need for background subtraction
  (ie, working with difference images)
M-class flare of 9-sep-02 
Dennis comment about using peak flux instead of centroid of 
  background-subtracted images (Schmahl, Hudson, Ji discussion)
Height of coronal source shows initial downward motion?
1-nov-04 Purple Mountain event
17-aug-02 BBSO data shows kernels approaching, again downward motion
  in early phase
26-jul-02 too compact to separate the kernels; but position again shows
  consistency with inward motions
Conclusion: hope to infer reconnection behavior